A Benchmark of Photobombing Removal

A Benchmark of Photobombing Removal

Presenter(s)

Vatsa Sanjay Patel

Comments

Presentation: 1:20-1:40 p.m., Jessie Hathcock Hall 180

Description

Photobombing occurs very often in photography. This causes trouble for the human subjects of the photographs. Thus, there is a valid requirement to eliminate photobombing from captured images. In this study, we do a comparative analysis of this topic. In order to accomplish this, we first collect a dataset of photographs including unwanted and distracting components that must be removed in order to detect photobombing. Then, we annotate the photobombed areas that have a tendency to be eliminated. The photobombed portions are then removed using various picture inpainting techniques. The photobombing removal results are compared to a manually created ground truth. Many performance measures are employed in our benchmark to compare the efficacy of various techniques. Experiments yield illuminating data that indicate the efficacy of image inpainting techniques for this specific issue.

Publication Date

4-19-2023

Project Designation

Graduate Research

Primary Advisor

Van Nguyen

Primary Advisor's Department

Computer Science

Keywords

Stander Symposium, College of Arts and Sciences

A Benchmark of Photobombing Removal

Share

COinS